RADDS "dry test" problems. (LCD menu)

Hi everybody,

I am planning to replace the existing controlles in my K8200 printer with a RADDS/RAPS128 combination.
For this I have a RAMPS 1.5 board with a genuine DUE and four RAPS128 mounted (in correct orientation, jumpers at 128 mikrosteps).
Vref is set to 0.9V for all of them.

I have compiled Marlin4DUE (latest from githup) with some adaptions (e.g. motor enable inverted).

I have NOT connected any hardware yet, just the DUE/RADDS with the normal LCD/encoder combination.

When I flash the firmware, the system gets power only over the programming port. It still starts and provides the menu. I can step through the menu with no problems.

But when I remove USB power and attach 24.2 V to the power supply (the one which powers steppers, NOT the heated bed, which stays unconnected), I have a different behaviour:
Almost everytime shortly after I switch to other menu entries (with a short beep) it jumps back to the main menu screen (with another single short beep).
If I switch to other menu entries very fast, some times it stays there (it jumps back less often when I am deeper in the menu tree), and then I can even correctly move a motor by using the menu (!).

I am not experienced with Marlin in combination with LCD menu (my K8200 does not have one).
So are there "normal" circumstances which makes the menu jump back to main menu all the time? There is no error shown, only something like "Marlin4DUE ready".

I would rather know if I can "trust" this controller before rewiring my whole printer. Any ideas? As mentioned: the menu works fine when power is supplied over USB only, and fails when powered with 24.2V, even if no motor is connected.

Any feedback is appreciated,

Best regards,
Michael

One additional note:

When I am deeper in the menu tree, it jumps back to main menu upwards level-by-level, not in one go.
It even jumps upwards, when the "focus" is already moved to other menue entries (i.e. not on the first entry, which would be used to go back "manually").
When I stay in main menu, there is no beep or false input.

Best regards,
Michael

Maybe some missing pullups. You could look in the init for the display. A pullup is set up by PULLUP(pin) and not WRITE(pin,1). This could be the issue. Anyhow, I heard about some issues with a not complete assembled RADDS. After assembling any compound (stepper, etc.) it magically stoped.

---

Triffid Hunter's Calibration Guide

--> X <-- Drill for new Monitor

Most important Gcode.

I couldn't find issues with the PULLUP so far (it seems like any pin defined as input has the pullup-definition as well), but my understanding of the code is still limited and incomplete.

But yesterday I might have found the problem. In my "dry test" there is still no mount or case around the PCB boards, and when I manipulate the menu, my fingers touch the back of the PCB. When I touch it more or less underneath the push buttons (the one below the reset), the menu steps back. I assume my fingers have a resistance of at least 50 or 100 Kiloohm, so a missing pullup is probably the reason.

Maybe there are some other problems (contacts?) as well; sometimes it mis-detects the SD card, and the menu reacted when I (more or less softly) dropped the board to the desk.

But when the LCD is untouched and not moved, the rest of the RADDS seams to work fine. So I'll try to wire it to my printer and see how stable it gets...

Thank you,
Michael.

BTW: Ich muss mich dafür entschuldigen, den ganzen Thread auf Englisch geführt zu haben. Ist wohl eine Berufskrankheit. Ich habe erst zu spät gemerkt, dass zumindest in diesem Teil des Forums in Deutsch geredet wird. Kommt immer seltener vor, dass ich über technische Belange in meiner Muttersprache kommuniziere. Armes Deutschland :-)
Sollte irgendjemand Verständnisprobleme haben, kann ich gerne relevante Details auf Deutsch wiederholen.

...Du Nase

---

Triffid Hunter's Calibration Guide

--> X <-- Drill for new Monitor

Most important Gcode.

Nach der Erkenntnis, dass meine dicken Finger den Back-Button hochohmig betätigt haben, habe ich mir die lcd-init nochmals angeschaut.
Tatsächlich ist der "btn_back" weder als input, noch als pullup definiert. Ein wenig Rumsuchen in anderen Dateien hat mich überhaupt erst darauf gebracht, dass es btn_back gibt.
Der Fix war dann einfach:

#if BTN_BACK > 0
SET_INPUT(BTN_BACK);
PULLUP(BTN_BACK,HIGH);
#endif

So funktioniert das LCD board samt encoder endlich störungsfrei.

Vielen Dank für die Hilfe,
Michael

PS.: Ich arbeite seit vielen beruflich mehr oder weniger ausschliesslich mit Java/SVN, daher ist mir GIT nicht ganz geheuer. Ich würde den Fix gerne zur Verfügung stellen, aber vielleicht fällt es einfacher, die vier Zeilen einzubauen, als mir zu erklären, wie ich dass sauber selbst "contribute" (git push? pull request?, ???).

Ah, das war mein Fehler. Ich pack das morgen rein. Danke!

---

Triffid Hunter's Calibration Guide

--> X <-- Drill for new Monitor

Most important Gcode.